Porosity prediction

Abstract

Examples of methods for predicting porosity are described herein. In some examples, a method includes predicting a height map. In some examples, the height map is of material for metal printing. In some examples, the method includes predicting a porosity of a precursor object. In some examples, predicting the porosity of the precursor object is based on the predicted height map.

The invention claimed is: 
     
         1 . A method, comprising:
 determining, by a processor coupled to a printer, a height map indicating one or more variations in a height of material in a build volume of material for metal printing based on one or more agent maps, wherein the one or more agent maps provide data indicating one or more locations for applying one or more agents, and wherein the printer is operable to manufacture an end object;   determining, by the processor, a porosity of a precursor object based on the determined height map by determining a porosity for each voxel of the height map;   determining by the processor, a predicted shape of the end object based on the porosity of the precursor object;   determining, by the processor, one or more predicted shape deformation based on one or more differences between the predicted shape of the end object and a corresponding 3D object model;   adjusting, by the processor, a 3D object model and one or more printing variables of the end object to reduce the one or more predicted shape deformations to a targeted amount; and   controlling, by the processor, the printer to manufacture the end object based at least on the adjusted 3D object model and the adjusted one or more printing variables.   
     
     
         2 . The method of  claim 1 , wherein determining the height map is based on a height prediction machine learning model. 
     
     
         3 . The method of  claim 1 , wherein determining the height map is further based on sensed height data or slice data. 
     
     
         4 . The method of  claim 1 , wherein determining the porosity is based on a porosity prediction machine learning model. 
     
     
         5 . The method of  claim 1 , wherein determining the predicted shape is based on a shape prediction machine learning model. 
     
     
         6 . The method of  claim 1 , further comprising determining an expected porosity of the precursor object based on the predicted shape. 
     
     
         7 . The method of  claim 6 , further comprising determining an expected height map based on the expected porosity. 
     
     
         8 . The method of  claim 7 , further comprising determining an expected slice based on the expected height map. 
     
     
         9 . The method of  claim 8 , further comprising performing online compensation based on the expected slice. 
     
     
         10 . The method of  claim 8 , further comprising performing offline compensation based on a set of expected slices.